The pharmaceutical, biotechnology and device industries fund most clinical trials. However, accountability in such trials is the subject of concern, due in part to evidence that such trials tend to favor the company's product, to worries about underreporting of adverse events, and to recent disclosures about ghostwriting. One major mechanism of accountability for industry-sponsored research is the presence of an academic Principal Investigator (PI), whom journals, clinicians and the public view as the trial's guarantor. However, no data are available regarding the roles that PIs play or the degree of control they exercise over major tasks and decisions. Consequently, it is difficult to know whether it is reasonable to view PIs as guarantors of their trials. Furthermore, there are no detailed published standards regarding the PI's role. The Aims of the present proposal are: 1) to articulate a detailed normative conception, based on major stakeholders' views, of the roles that overall PIs in multicenter randomized controlled trials (RCTs) should play; 2) to describe attitudes among PIs of multicenter RCTs published in major journals towards the appropriate roles and responsibilities of the PI; 3) to describe the actual roles that PIs played in particular trials; and 4) to test the hypothesis that PIs in industry-sponsored multicenter RCTs play different roles than do PIs in non-industry sponsored trials. To achieve these Aims, we will conduct 6 Delphi panels with representatives of diverse stakeholder groups (NIH clinical research leaders; clinical trial cooperative-group and network chairs; PIs; industry clinical research leaders; journal editors; FDA officials). These panels will be followed by parallel surveys of PIs of multicenter RCTs published in major general, oncology, cardiovascular and psychiatry journals to assess both their normative views of the PI role (survey 1) and their actual practices as PIs (survey 2). These panels and surveys will provide the first detailed description, taking into account the views of key stakeholder groups, of the responsibilities and expectations of overall PIs in multicenter trials. In addition, they will provide data regarding the actual roles and responsibilities of PIs in a diverse set of trials, providing the first systematic answers to questions about PI accountability in clinical research.